Single Pane of Glass view for Application Network KPIs using Dashboards
In my previous blog post, Dashboards in Aria Networks v6.9, we got introduced to a new framework that enabled users to create, view and manage dashboards within their setup.
One of the key asks from users is to provide them with a single view showing all network performance metrics for their applications and the health of the underlying infrastructure that hosts their hybrid applications.
In this blog, we will go step-by-step over an Application Dashboard showing various network performance KPIs and insights without the need for multiple navigations.
Some of the key customer requirements are as follows:
- As an Application owner, I need a dashboard that shows me the overall Topology map of my hybrid cloud application, inter-dependencies, and all the alerts and degraded flows showing health issues within the underlying infrastructure components.
- As a Kubernetes administrator, I need a dashboard showing the topology map of my microservice-based applications along with inter and intra-network traffic within pods serving my k8s-based service, and a topology map showing the health of overlay and underlay gears with end-to-end connectivity.
- Show me the Busiest VM flows and sessions to optimize the application further and balance out workloads across other entities within a Tier.
- As a Database administrator, I want to see the overall traffic rate for a database cluster and if traffic is properly load-balanced and being served by the backend DB nodes.
- As an infrastructure administrator, I would be interested to see if there are any abnormal flows for Shared services (Ex: DNS, NTP) and if it is impacting any application performance.
Dashboards become the landing page for application teams showing complete networking insight into their Applications and the health of the backend infrastructure hosting these services while the Guided Network Troubleshooting framework helps to quickly identify performance bottlenecks within the user environment thus improving the overall MTTI (mean time to identify) and MTTR (mean time to resolve) KPIs.
Application Network Dashboards
Creating Dashboards is easy and follows the same procedure of widget pinning as seen in Pinboards. Custom search query results can also be pinned to a Dashboard. These dashboards can be shared with users and other teams and the URL can be bookmarked to launch them quickly.
Here, we look at one of the Application Dashboards.
The Summary page shows a quick overview of open alerts for your application, and traffic visibility (IN/OUT), it shows the total count of flows and unprotected flows where NSX firewall rules need to be applied. Intents show Flow health and utilization violations. Also, a quick VM inventory and flows and user traffic breakup by country/region.
Application Topology shows the overall topology map of the hybrid application and the health of every tier and backbone network, the interdependent application tiers, and flows traffic between tiers (Ex: web tier talking to DB tier and the internet), the switches, routers, load balancers with backend connectivity. The interdependency with Shared services and other applications and overall, Tier health.
Alerts show all correlated alerts and underlying infrastructure alerts impacting the overall application. Use the “Troubleshoot” button to launch into the Guided Network troubleshooting framework for further troubleshooting and identifying potential root causes.
The Members Manager widget shows the overall health of the providers: vSphere vCenters, NSX, VMC SDDC, and account status for Public cloud integrations (Ex: AWS, Azure).
Outlier Analysis shows total traffic breakup as well as sessions within the application Tier ( Ex: How many requests have been handled by individual web servers within the web tier).
VM Metrics show the network traffic KPIs and the vSphere properties (VM IPAddress, VLANs, Disks, datastore, ESXi host, etc) and Troubleshoot button to launch into Guided network troubleshooting framework for identifying potential root causes.
The external Services widget shows external internet services accessed in the last 24 hours.
Using top talkers, find the busiest VMs within the environment based on flow count, opened sessions, and total traffic size allowing optimization opportunities within your application.
Identifies abnormal flows based on total TCP round trip time (RTT). This allows users to detect if connections are timing out within the application or to other shared services accessed.
Users can use the Micro-segment wheel view to analyze the traffic flow patterns for better security planning. The blue lines denote the outgoing flows, the yellow lines denote incoming flows and the green lines denote bidirectional flows.
Similarly, you can build custom dashboards for infrastructure-specific ones using Pinning and Search queries ( NSX-T health, ESXi host nic metrics, etc )
Below are a few more examples of Network Dashboards:
Kubernetes micro-service Application Dashboard
This post should give you a good understanding of how Custom Dashboards can provide all the necessary information for an application and its underlying dependent entities on a single page without the need for multiple navigations. Using Dashboards, we can monitor and get end-to-end visibility for any application with the backbone networking infrastructure and KPIs.
TechZone videos on Aria Operations for Networks